Organizational Regularities in Recurrent Neural Networks

TL;DR

This paper explores how organizational regularities in weight matrices of recurrent neural networks influence their dynamics and performance, revealing that certain regularities like modularity improve task accuracy, while others like reciprocity may hinder it.

Contribution

It introduces a systematic analysis of biological-like regularities in RNN weights and their effects on dynamics and computational performance.

Findings

Dale's monopolarity and modularity improve task accuracy.

Reciprocity tends to reduce performance by causing early saturation.

Regularities can be tuned to control network dynamics.

Abstract

Previous work has shown that the dynamical regime of Recurrent Neural Networks (RNNs) - ranging from oscillatory to chaotic and fixpoint behavior - can be controlled by the global distribution of weights in connection matrices with statistically independent elements. However, it remains unclear how network dynamics respond to organizational regularities in the weight matrix, as often observed in biological neural networks. Here, we investigate three such regularities: (1) monopolar output weights per neuron, in accordance with Dale's principle, (2) reciprocal symmetry between neuron pairs, as in Hopfield networks, and (3) modular structure, where strongly connected blocks are embedded in a background of weaker connectivity. We construct weight matrices in which the strength of each regularity can be continuously tuned via control parameters, and analyze how key dynamical signatures of the RNN evolve as a function of these parameters. Moreover, using the RNN for actual information processing in a reservoir computing framework, we study how each regularity affects performance. We find that Dale monopolarity and modularity significantly enhance task accuracy, while Hopfield reciprocity tends to reduce it by promoting early saturation, limiting reservoir flexibility.